E-mail: The Art of the Append
By and large, people have come to accept the delivery of direct mail in their postal mail box. The same is not true of e-mail. Perhaps because people set up their own addresses and pay for their account service, they are more guarded about who and what is allowed into their e-mail inbox.
This viewpoint is what makes the practice of e-mail appendingwhere you match up customer names with e-mail addresses compiled by third party vendorscontroversial.
"The privacy zealots will say that any message the recipient did not give the marketer permission to send is spam," explains Reggie Brady, proprietor of Reggie Brady Marketing Solutions, a direct and e-mail marketing consultancy.
But everyone's definition of spam is different, Brady points out, and it's not a given that your customers, who already have established some kind of relationship with you, won't want to hear from you via this channel.
You can reduce the likelihood of upsetting customers, says Kim MacPherson, CEO, creative director and founder of Inbox Interactive, an e-mail marketing services company, by appending the right way: determine the most appropriate customers to append; choose a reputable appending service; analyze the match criteria/results; and create relevant contact strategies.
Who to Append
A good measure of customers' willingness to hear from you via e-mail is the recency of their last contact with your company, says Brady. "The more recent the activity, the more likely they are to be responsive."
Most marketers, Brady notes, will benefit from segmenting their house file into three-month, six-month and nine-month buyers, and testing each before rollout; based on successful results from these campaigns, marketers can work their way deeper into the file.
But don't get append happy. In its set of best practice guidelines for e-mail appending, The Association for Interactive Marketing's Council for Responsible E-mail (the C.R.E.) recommends appending customers that have purchased within the last two years, says Scott Shamberg, director of product development for YesMail Append, at e-mail marketing solutions company YesMail. Any older, and the existence of a valid business relationship comes into question.
Recency is not the only consideration, however. Since not all sales are simple, one-step transactions, MacPherson points out that you need to evaluate your definition of a "customer." For example, people who sign up for a trial offer but have not yet converted to paid status might be good prospects for e-mail follow-up.
The same is true of business-to-business leads, which may be involved in a longer sales cycle before conversion. In fact, says Brady, the b-to-b sales process lends itself very well to e-mail communication.
She adds that it's best to start with more qualified leads, and keep in mind that the match rate on the append will be very lowbetween 7 percent and 10 percent.
How it Works
"In its simplest form, a customer house file is matched against a compiled database of permission names," explains Shamberg.
The keys to a strong matching process are: the cleanliness of both the house file and the compiled database; the ways in which the permission names were captured; how often the file is updated, and when it was done last; and the level of match desired.
"A match can be performed at many levels, but an individual level match will produce the most accurate information," says Shamberg.
Brady explains that match rates can range from 10 percent to 30 percent or 40 percent. But, she notes that match rates higher than 20 percent likely point to a household match and not an individual match. This means the e-mail address belongs to someone in the same household as the customerwhich could be a teenager who signed up to get a free sample of shampoo, and not the father, who is the primary target.
Generally, says Brady, the compiled files used by append services tend to be a mixture of people who have registered to receive free product samples or enter contests. Another source is that of permission names gathered by third-party vendors.
Reviewing the Match
The great thing about appending data is that you only pay for matches that fit the criteria you and the vendor agreed upon.
But you will pay for the names that opt out from receiving e-mail messages from you or that open your e-mail but don't respond to its opt-in invitation, says Brady.
While Shamberg states that these e-mail addresses will not be shared with clients, Brady advises marketers to have the append service tag the records that should not be contacted to prevent future privacy trespasses.
Shamberg adds that many append vendors also can perform an e-mail change of address, or ECOA, to boost the number of appended records.
TIP: Once the match is complete, look to see where the hits come from, Brady encourages. You're looking for the percentage of primary addresses, i.e., from AOL or Earthlink, versus. secondary addresses, such as those with Yahoo and HotMail accounts. Secondary accounts are often used by people as throw-away addresses that are set up for a particular need, MacPherson explains, and then rarely checked againwhether the person leaves them open or not.
Brady advises marketers to consider a separate contract strategy for these names, so they don't skew results for your overall append analysis.
After the Match
While other append processes typically end once the data has been merged, the e-mail append service usually handles the first e-mail contact. In fact, it's often a "contractual agreement based on an e-mail service provider's
e-mail append agreement with existing sources or contributors and their respective privacy policies," explains Al DiGuido, CEO of e-mail communications firm Bigfoot Interactive.
Additionally, a few benefits arise from this arrangement. For one, says DiGuido, "the e-mail append service should be white-listed with all the major ISPs and Web-based e-mail service providers to allow for the smooth and efficient delivery of an e-mail append effort."
DiGuido cautions marketers to be highly skeptical of e-mail append vendors that are reluctant to drop the first e-maila signal that they might have ISP relationship problems and/or a possible deficiency in the matching process, both of which spell trouble for delivery and complaint rates.
From a technical perspective, e-mail append vendors often can provide marketers with a higher quality deployment engine, he says. "For example, a top-tier append provider can offer the ability to track opens, serve HTML to AOL 6.0 [users], and provide more detailed reporting."
Another benefit is the knowledge of privacy benchmarks, established industry policies, as well as state and federal data protection laws that vendors often bring to the table, thus protecting more novice marketers.
A final benefit to having the append vendor handle deployment is the work in dealing with bounceswhich the vendor needs to know about to update its fileand opt outs, says Brady.
It should take about 10 days, she adds, for the deployment to be complete and the bulk of response to come in, at which point the vendor will return your file with only the valid e-mail addresses appended and flagged.
Altogether, match and deployment should run about $100/M to $450/M, depending on the volume and number of match records, says DiGuido.
The First E-mail Contact
Just as it's hard to get consensus on whether it's ethical to append e-mail addresses collected by third parties, opinions are mixed on whether the first e-mail contact should ask recipients to opt out or opt in.
"The C.R.E. has established the permission level for e-mail append as opt-out," Shamberg states. He adds that YesMail has done some opt-out mailings for clients, with mixed results. "Certainly a customer that opts-in to ongoing communication can be more active. However, when looking at e-mail append, the important metric to look at on the [first contact] is open rate. On average, we see open rates of 50 percent to 60 percent, meaning that over half of the recipients look at the message and choose not to opt-out."
While this can be true, MacPherson points out that people often don't opt out of e-mails that seek permission to contact them because of previous experiences with spammers who use opt-out replies to validate the e-mail address.
While an opt-in only message will definitely reduce the number of customers who give permission to use their e-mail address, MacPherson points out that you can do no harm with this approach. She adds that you also will be more sure of the names that you do add to your database.
How many names might you get on an opt-in message? Brady estimates that the maximum response you will get is 10 percent; if you choose the opt-out approach, the rate of opt-out by appended names is about 2 percent.
Should you choose opt-out for the first message, Brady explains, be sure the opt-out is immediately visible when the e-mail is opened and the instructions on how to opt-out are simple and crystal clear.
E-mail experts agree that the first contact should be more about gaining permission to establish an e-mail relationship than about making a sale. Since you are trying to convince these customers that there is a benefit to receiving regular e-mail communications from your company, MacPherson says it's perfectly fine to include some promotional content that tells customers what's in it for them.
"Certain verticals, such as retail, can include a soft or hard offer in the initial mailing as an example of the benefits they can get used to. However, this should be done under the umbrella of permission," Shamberg offers.
To track the results of your append carefully, it is not advisable to treat appended addresses the same way you do customer records with e-mail addresses you collected yourself.
Brady urges marketers who deployed an opt-out approach to carefully move these names into the regular messaging mix slowly, since they may be more sensitive to frequent contact. It could take as much as two to three months until these names are ready for migration.
One way to chart the progress of your e-mail relationship is to watch your e-mail links, noting which customers open e-mails and follow through with inquiry or transaction behaviors.
No response to the first few e-mails is a good indicator that something's amiss, whether it be a secondary address that doesn't get checked, a household-level match or merely no interest in your offer. Rather than continuing to shove a square peg through a round hole, Brady advises marketers to alter their message strategy to find the reason for the lack of response.
For example, she says, you could try special offers, ask recipients if they will share their primary e-mail address with you, or offer to alter the frequency or content of your messages to better suit their preferences.
For marketers that used an opt-in approach, no response means no future contactat least not in the near future. MacPherson reminds marketers to flag non-responders to the activation e-mail as "no e-mail" records only for the append effort. That way you don't preclude these customers from future contact if they share their e-mail addresses with you voluntarily.
As previously noted, it's not likely that you'll append addresses for all your incomplete records in one pass. But is there any benefit in going to a new append service or even going back to the same vendor for another try?
"The append process and timing are dependent on many variables, including how fast the customer list and existing offline database is growing and the growth of an e-mail append source file," says DiGuido. "Typically, most clients who have strong experiences with e-mail append set up quarterly updates."
One caveat, MacPherson notes, is that of re-contacting customers and prospects from a prior append attempt who requested not to receive e-mail from you. To reduce this risk, MacPherson stresses the need for marketers to share this information with the append service on every job.
Still, she recognizes that a change in life circumstances or a great offer might affect a customer's willingness to communicate via e-mail. She recommends testing a re-contact strategy in small segments, starting with those names with the most recent activity.